270 



HANDBOOK OF PHYSIOLOGY ^^ CIRCULATION I 



repolarization in any voltage range in response to 

 small (fig. 2 1 ^4) or large (fig. 21 5) hyperpolarizing 

 current pulses. Regenerative responses would be 

 indicated by inflections in the voltage-time curves 

 wliereas, in the records, these are always concave 

 toward the zero current 8 curve. 



If the conclusion is correct that there is no threshold 

 for repolarization in the voltage range at the end 

 of the second phase ( — 20 to —40 mv), most existent 

 hypotheses of repolarization in cardiac muscle (4, 

 24, 68, 106, 130, 131) are apparently incorrect. 

 The alternative is to suppose that tlie sudden changes 

 in g at the termination of the plateau are purely 

 time-dependent or only slightly voltage-dependent. 

 Shanes (106) suggests a mechanism of this sort, Init it 

 is difficuh to envision a mechanism which can 

 rhythmically increa.se and then decrease gK but 

 which is not voltage-dependent. Another possibility 

 is to suppose that g's do vary with voltage but so 

 sluggishly that the changes in £ produced by applied 

 currents are too brief to change the g's very much. 

 This possibility is contradicted by VVeidmann's 

 finding (125) that sufliciently large hyperpolarizing 

 pulses early in the plateau can initiate an early 

 repolarization, (cf. 28). This repolarization occurs 

 only if S is maintained more negative than the 

 resting potential for a period of time (cf. fig. 2ifi, 

 where equally large voltage changes scarcely affect 

 repolarization). The fact that maintenance of the 

 voltage around 8r may markedly alter the ionic cur- 

 rent flow after the end of the current pulse indicates 

 that there are voltage-dependent g's at some times 

 and some voltages. 



The existence in the heart of a threshold for early 

 repolarization does not contradict the above con- 

 clusions about the absence of a threshold at the 

 voltage of the second phase since the actual threshold 

 is more negative than 8r. This "threshold" may be 

 similar to the induced rep(jlarization of squid axon 

 (77) or frog a.xon (112) where a large repolarizing 

 current may terminate activity early. This abolition 

 response probablv can occur when there is only 

 one stable point or when all three equilibrium points 

 are still present. In the former case no threshold is 

 involved; any current would simply accelerate the 

 repolarization process. However, if there are three 

 equilibrium points, the early repolarization would 

 have a threshold current; smaller currents would 

 have little lasting effect on 8. The available evidence 

 indicates that there is no true threshold for repolari- 

 zation in cardiac muscle if [Ca++]o is normal. One of 

 Weidmann's records (125, fig. 5) shows a current 



applied at the start of the plateau carrying the voltage 

 below resting value and initiating early repolariza- 

 tion. A current just insufficient to initiate early 

 repolarization activates enough gxa that, on cessa- 

 tion of the current, 8 is in a region of negative G and 

 "anodal break" excitation occurs (fig. 21 B, lower 

 trace). A similar recording (125, fig. 6) shows a 

 threshold at about 45 mv above maximum diastolic 

 potential. However, in these records the current 

 was applied during the middle of the plateau and 

 the recording electrode was 4.2 mm from the current 

 electrode. Hence the voltage change at the current 

 electrode was se\eral times bigger. 



Cranefield & Hoffman (28) found that lowering 

 the [Ca++]o to one-fourth of normal decrea.sed the 

 amount of current recjuired to induce repolarization. 

 In this medium, the soltage-time cur\e during 

 hyperpolarizing current flow was slighth', but defi- 

 nitely, inflected. This inflection indicates a regenera- 

 tive process with a threshold for repolarization at 

 about the knee of the normal repolarization curx'e 

 (28, fig. 10). VVeidmann (128) found that raising 

 [Ca''~'"]o shifted the curve relating 8<i to initial voltage 

 (fig. 17) toward zero voltage, i.e., the amoimt of 

 gN:i activation at any particular voltage is increased. 

 These findings accord with Frankenhaeuser & 

 Hodgkin's (45) voltage clamp studies of the effects 

 of [Ca++]o variations on .squid axon. In this study 

 the effects of a rise in [Ca++]o on the kinetics of 

 gxa changes were almost identical with the effects of 

 hyperpolarization (cf. 77). A reduction in [Ca++]o 

 has the opposite effects in squid axon and probably 

 in the heart also. It follows that the inactivation 

 attendant on a reduction in [C'.a++]o will reduce the 

 probability of anodal break excitation. Nexerthe- 

 less, it is difficult on the basis of this information to 

 deduce the mechanisms responsible for the shift 

 from a nonthreshold- to a threshold-type behavior 

 in the early repolarization phenomenon when 

 [Ca++]o is lowered. One possibility is that a reduction 

 in [Ca++]o reveals an underlying regeneratix-e re- 

 polarization process which is ordinarily counter- 

 balanced by gxa actixation. 



The abolition of the action potential by hyper- 

 polarization can be explained, but the possibility 

 that this early repolarization may propagate even 

 in normal [Ca++]o is puzzling (28, 125). The spatial 

 voltage gradient in a propagating early repolarization 

 xvould be small and there xvould be little local current 

 floxv. If [Ca++]o is loxv and there is a threshold, then 

 the process could propagate at a loxs- speed. However, 

 the absence of an energx-vielding threshold or non- 



